[gmx-users] Re: 1-4 interaction between atoms in PBC mode

[Darrell Koskinen]

Hi Omer,
The atoms in question are part of a graphene sheet and are located at 
two opposite ends of the box. They should not move as they are bound 
inside the graphene sheet. The initial separation between these atoms is 
19.678 nm, which is the same distance as in the warning message. So the 
position of these two atoms apparently does not change from the 
beginning of the simulation until the point at which the warning occurs.

Darrell
> Date: Mon, 21 Sep 2009 10:21:20 +0300
> From: Omer Markovitch <omermar at gmail.com>
> Subject: Re: [gmx-users] 1-4 interaction between atoms in PBC mode
> To: Discussion list for GROMACS users <gmx-users at gromacs.org>
> Message-ID:
> 	<8c3f24a40909210021p685a1e7am83c1fac75e626bf2 at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
>   
>> I assume that atoms from opposite ends of the simulation box are being seen
>> by GROMACS as being 1-4 neighbors as a result of the PBC,
>>     
>
> Are you sure? have you looked at the last "good" frame before this message?
> GMX treats PBC as it should during MD neighbour search, so maybe this is a
> pair that was initially close and interacting, but later moved away from
> each other.
>
>
>   
>> but the shortest distance between the atoms is not being observed. Is this
>> the problem? Is this problem resulting in the system exploding? And, how do
>> I resolve this problem?
>>
>>     
> If you are sure that everything else is ok, then extend the table extension.
> Two other things to work around this are either reduce timestep or check
> that your definition of the box is as it should.
> --Omer.
>
>